Project Rescue

Your build is stuck.
I get it shipped.

For founders whose AI-built MVP is breaking at scale, whose freelancer ghosted, or whose project stalled mid-launch. Scoped recovery work with clear deliverables. No retainers required, no open-ended consulting.

Book a Production Readiness Review, $2,500 →What's actually broken?
$2,500 · 7 days · ranked report you keep regardless of next steps
23 findings7 P03 data-loss risks
This is for you if…

You shipped something that worked, then it stopped working at the wrong moment.

  • You shipped fast with Cursor, Lovable, Claude Code, or a contract dev, and now real users are breaking it
  • Your codebase has issues you can't pin down, auth boundaries, scaling, data integrity, security
  • An enterprise deal is stalled because your tech can't pass a pentest, DPA review, or due diligence
  • A freelancer disappeared mid-build and you have to make their work production-ready
  • A migration, refactor, or platform-switch stalled and the team can't get unstuck

If two or more of those describe you, the Review is built for your situation.

What rescue actually looks like

Four patterns I keep seeing.

No two rescues are identical. But the failure modes rhyme. Here are four shapes the work usually takes, anonymized, condensed, real.

Pentest finding
"Found a dashboard during a pentest with no row-level security. Changed one URL parameter and saw every customer's data."
B2B SaaS · 11-day fix · enterprise deal closed two weeks later
Enterprise blocker
"6-month-old MVP, 3,400 lines of code in one file. First enterprise pilot asked for SSO. Rewrote auth in 11 days."
AI-built MVP · 1 founder, 0 engineers · pilot → contract
Scaling failure
"Staging passed, prod crashed under 200 concurrent users. Singleton state in a stateless service. One-week fix."
Series A B2C · launch week · zero further outages
Acquisition DD
"Acquirer's tech due diligence flagged no observability, no runbooks, no CI. Fixed all three in 3 weeks. Deal closed."
Bootstrapped acqui-target · 3-week sprint · 7-figure exit
How the work happens

The pipeline.

Diagnostic first, always. I won't quote a Rescue Sprint until I've read the codebase. The Production Readiness Review is the gate, and you walk away with the report whether or not we proceed.

01

Triage

30-min call. What's broken, what's at stake, what's the blocker. Decide if rescue is the right tier or if you need something else.

02

Diagnose

Full read-through of the codebase, infra, and recent incidents. I find what's actually wrong, not what you think is wrong.

03

Prioritize

Ranked report. Severity, blast radius, time-to-fix, dependencies. Some items might not need fixing. Honest call.

04

Scope

Fixed price, fixed scope on the items that matter. If the rescue uncovers more than expected, I tell you within 3 days, before scope creeps.

05

Rescue

Two to four weeks of focused work. The pipeline runs; I review every diff. Real fixes on the real codebase, not advice.

06

Hand off

Written handoff with what changed, why, and how to maintain it. Repo stays yours, history stays clean, you walk away whole.

Honest scope or no scope. That's the whole agreement.

What's included

Two tiers. You start with the Review.

Pricing is fixed and upfront. The Review stands alone. The Sprint only runs after a Review.

Step 1 · always start here

Production Readiness Review

$2,5007 days · fixed price · ranked report delivered
  • Full code review against production-readiness criteria (auth, data integrity, scaling, observability, security, CI)
  • Infrastructure review: hosting, secrets, deploys, backups, monitoring
  • Ranked report: severity, blast radius, fix estimate, dependencies
  • 1-hour walkthrough call once delivered
  • The report is yours. Use it with me, with your team, or with someone else. No strings.
Book the Review →
Step 2 · only after a Review

Rescue Sprint

$6,000 – $9,0002–4 weeks · fixed scope · fixed price
  • Fix the top-priority items from the Review
  • Fixed-scope, fixed-price. No hourly billing. No "we'll see how it goes."
  • The pipeline executes; I review every diff. Same model as AI Dev Pipeline, scoped to recovery work.
  • Daily async update on progress. Weekly call.
  • Written handoff at the end: what changed, why, how to maintain it
Learn more →
Proof

Diagnostic depth. Verifiable artifacts.

Some rescues are under NDA. Where the work is mine to show, here's what the diagnostic depth looks like.

Public · Open source

The same diagnostic depth applied to my own engineering toolchain, every day.

475 tests, 21 files updated daily, every commit reviewed. The same eyes that find the row-level-security bug in a client's auth boundary are the eyes that maintain my own production system. The diagnostic instinct comes from doing this work continuously, not from a one-off audit.

475automated tests, all passing
21files updated per day, on average
100%commits human-reviewed before merge
See the dogfood case →
Pattern · Auth

The dashboard with no RLS.

Most AI-built MVPs have it. Found in 7 of the last 11 reviews. Always fixable, always cheap to find, always devastating if a customer finds it first.

Pattern · Scale

Singleton state in a stateless service.

Passes locally, passes staging, breaks at 200 concurrent. Found in 4 of the last 11 reviews. One-week fix every time.

FAQ

Straight answers.

The questions every founder asks before they admit something's broken.

That's exactly what the Production Readiness Review is for. $2,500, 7 days, you walk away with a ranked list of risks and fix estimates whether or not we work together. Most clients use it as decision support, not as a contract.
Yes for the Review. Rescue Sprints typically start within a week of the Review wrapping. If your situation is genuinely time-critical (board meeting, pentest deadline), say so on the triage call and I'll tell you honestly whether we can hit it.
I tell you within the first 3 days, before scope creeps. We either re-scope honestly or stop. I'd rather lose the engagement than carry it under bad terms. You get the diagnostic regardless.
No. If your prod is on fire right now, this is not the right tier. I do scoped recovery, not 24/7 on-call. If you need a war-room engineer, I'll refer you to one.
You do. Code stays in your repo, in your name, with full handoff docs. I don't lock you in. The whole point is to leave you self-sufficient or with a clean handoff to your next hire.
Web stacks: Next/Node, Rails, Django, Go, Phoenix. Mobile and embedded, I refer out. If you're on a wildly unusual stack, mention it on the triage call.
Sometimes. If the issue is "we have a backlog of fixes we can't get to," the pipeline is right. If it's "we don't know what's wrong," you need rescue first. The Review is how we figure out which.
I'll tell you, with a cost and timeline. Sometimes the answer is a Rescue Sprint that keeps things alive while you plan a clean rebuild. Sometimes the right call is to start over, see MVP Build. I won't recommend a rewrite to bill more hours.

Two honest doors.
The Review is the same either way.

Book a Production Readiness Review, $2,500 →Tell me what's broken
7 days · ranked report · yours regardless of next steps60-second form · I tell you if I can help, or who can

If you'd rather start over than rescue, see MVP Build. If you need ongoing partnership instead of recovery, see Fractional CTO.

Matt